Re: [Rfced-future] Model proposal

[Brian E Carpenter <[e-mail]>]

On 08-Jul-20 12:03, Larry Masinter wrote:
> ... My concern is that
>> this needs to be facilitated and informed by someone with relevant
>> knowledge and experience in the editing/publishing/library/archival world
>> and that is not us
> 
> 
> You keep on making this assertion that the skills of editing, publishing, archival world "is not us".  I think being versed in the engineering considerations of those properties SHOULD be "us".

Possibly, but the fact is that I learnt a lot more about those things from the only RSE with those skills than I did from any of her predecessors back to Jon Postel. So I am not making this up or theorising; I'm saying this is my lived experience. Heather knew stuff that none of {Jon, Joyce, Bob, Glenn, Olaf} knew.

I've worked with physicists and software engineers all my life, and we tend to be pretty arrogant about our ability to take strategic decisions in areas we know little of.

> The reason we have an archival series of specifications is not that we want to build a memorial Ozymandias altar to our golden words, but because these are the archival records of how things worked. It's a serious problem when buildings and bridges last longer than the documents that describe them.   How else can you make all of these paper-based archival solutions into all electronic worlds?

Indeed. Another whole area of specialization that Heather knew a lot about. No, she didn't know a whole lot about protocol engineering.

> But all of this is an engineering problem, within the range of what I'd expect an engineer to deal with.

Well, it's a full time specialization as far as I can tell from a quick look at IEEE Xplore and the ACM Digital Library. Not something we can do during coffee breaks from protocol design.

So do we need an expert consultant? Yes. Do we have to stick rigidly to the traditional RSE model? No.

Regards
     Brian

> We're spending a lot of time engineering a person-to-person protocol. The unique parts of IETF is that we treat person-to-person protocols at layer 7.5 with checks and balances, without any work on writing down the Security Considerations.